Square Pegs

Thanks Patrick

There are a few changes I am going to make but your advice is right on. I have been trying to decide on a direction to take and your experience is invaluable.

I am going to sign off for the weekend.

Thanks again

Stan

Thanks! Art and Earl know more about these older boxes than I do though. I've only been at this for a bit over a decade.
 
Stan,
That horn was not created to produce a diffraction effect, that was just the way that was chosen to load the driver at what looks like 50% of the cone area. It could be done with a circular opening but the way they designed the horn it was just simpler to just cut the cone area in half by a straight edge. I would think at high excursion that the surround would hit the back of that plywood area in front of the cone if there was not some sort of spacer in front of the bass speaker.
 
I would think at high excursion that the surround would hit the back of that plywood area in front of the cone if there was not some sort of spacer in front of the bass speaker.
The EV15BW and 15BWK used in the Centurion were designed for horn loading and fitted with a gasket deep enough for the expected excursion which was probably less than 3.3 mm, about 1/8" one way.

Other speakers may need an additional gasket.

I would agree with Patrick's suggestion regarding replacing the mid and HF with a coax, the original folded PA mid horn is more appropriate for carnival midway rides than home use, though the T-35 tweeters are not bad, if the diaphragm is properly shimmed and centered.

Those changes would require a crossover designed for the components, and delay of the mid/high to the LF horn path time of flight would be an improvement.
 

Attachments

  • Centurion Drivers.png
    Centurion Drivers.png
    334.3 KB · Views: 851
weltersys,
I brought up the spacer because I have had to do that on a large double 18" driven horn with high excursion JBL drivers. The spacer were substantial in that instance but they were for PA use and were driven by some high powered Crest class D amplifiers. Let's just say they made massive amounts of bass.
 
Thanks everyone.

I didn't mean to hijack your thread but as I said, my Centurion is a mess in oh so many ways. Not only was it poorly constructed, (sorry dad) but as you guys have pointed out, there are some basic design flaws with the Centurion.

I started asking about running a coax through a folded horn but quickly realized just how bad an idea that was, but I am still not convinced that a 16" Paraline isn't a good way to couple a 16” driver to a horn. Looking at the picture in #673 I don’t know where the throat of the horn is. Is it the whole semi-circular cut out or is it the rectangle formed by the straight edge as it is projected to the other side of the horn?

The reason I finally decided to jump into this thread in is it occurs to me that a Paraline will deliver 100% of the sonic energy efficiently to the throat of the horn, which is at the apex of the horn. In comparison, the semi-circular throat seems like it will deliver a mushy sound at best.

A little bit more about myself: I am an engineer by training and temperament. I am very good at visualizing things in 3D. I understand the physics behind optical systems and how lenses work. I usually don’t plan out the things I build because I work out the plan in my head, although I do make sketches to work out dimensions. There was a decade long section of my career that I designed circuitry, test fixtures, tooling, and airplane parts in CAD.

Saying that is to support my contention that a Paraline is good method to couple my woofer to my horn. I am going to use a Great Plains Audio 515-8LF and I don’t think I will need a spacer as they are designed for rear mount.

Patrick, you have said that you have been doing this for only ten years. Well, I have been thinking of doing this for only ten months, but I have so much confidence in what I see that I am going build a Paraline to couple my woofer to my horn.

The 515-8LF is an inch bigger in diameter than the Jensen that my dad used so I have to replace the plywood anyway, and while I am at it I will turn the LF horn into a properly shaped horn, not one with myriad sharp edges, constrictions, and contractions.

I realize I need my own thread for this rebuild but I am wondering if you guys are interested in seeing what I come up with. After all, out of the mouths of babes frequently comes wisdom, and if nothing else, I am certainly a newbie.

Stan
 
e2dedf59-c0b8-4918-a581-04b443238801.jpg


In Houston, there's a sculpture which acts a bit like the Paraline does. Here's how the sculpture works:

1) You sit down and say something
2) The sound from your voice radiates spherically
3) A segment of that sphere is collected and reflected by the marble parabolic reflector
4) The sound is radiated to a parabolic reflector about 20 feet away
5) The same process happens in reverse
6) And the person at the opposite sculpture can hear you as if you were standing right there.

It's a trippy effect; if you move your head two inches the effect goes away, but line it right up and it sounds like they're right in front of you.


An externally hosted image should be here but it was not working when we last tested it.


It occured to me that you could make a three dimensional curved array using a satellite dish. They're easy to find; on my local CL there's a few dozen for sale. As cheap as $10.

Might be an interesting project, particularly for situations where you don't want to bug people in the room, or your neighbors. For instance I work in front of a computer all day, and I like to listen to podcasts. Sometimes I won't listen to certain recordings because I don't want the kids to hear them. (cough cough 'the Opie and Anthony show'). A converging array will be loud at one point, and quiet everywhere else (due to comb filtering.)

[/font]
 
I have been thinking about the paraline for a bit, wondering if it would make it possible to somehow make a domestically friendly line array speaker without going floor-to-ceiling.

One thing that struck me is that although the path lengths for in the "standard" paraline that creates a flat wavefront are equal, the output along the line is not. Perhaps that is a good thing - perhaps not in all cases. But it seems to me that it would be interesting to have a device with equal output along the paraline along with a flat wavefront.

Here is a standard paraline (one quarter of it anyway), split into four 22.5 degree slices:

attachment.php


You can see how when you fold the lines back, you get the same path lengths for each segment.

However, each slice of the pie gets a different horn profile. This is easier to see if we mirror the "folded back" sections:

attachment.php


Different lengths of "expanding (conical?) and tube-like" expansions for each section.

I am wondering if this could be one of the reasons for the uneven response seen by people who have tried building these.

So I tried to draw a variant that would have equal expansion from the throat to the mouth. Through trial and error it seems like the only way to do this, is to make the mouth smaller, using only half of the area available radiating away from the driver. In the example below I´m using three slices with 12% angle. You can see how the angle continues as that section is reflected back towards the line opening.

attachment.php


Here´s how the first two layers would look like:
attachment.php


Third layer added:
attachment.php


Fourth layer added:
attachment.php


Final layer added:
attachment.php


Now, I didn´t place any reflector or anything there yet, but this will have to do for illustration. :)

Anybody care to give critical but friendly feedback on these thoughts?
 

Attachments

  • paraline1.png
    paraline1.png
    95.2 KB · Views: 721
  • paraline2.png
    paraline2.png
    94.5 KB · Views: 729
  • paraline3.png
    paraline3.png
    83.7 KB · Views: 719
  • paraline4.png
    paraline4.png
    117.4 KB · Views: 709
  • paraline5.png
    paraline5.png
    84.9 KB · Views: 707
  • paraline6.png
    paraline6.png
    107.7 KB · Views: 714
  • paraline7.png
    paraline7.png
    80.4 KB · Views: 708
Yes, I have seen the video. It seems easy enough to make the curve you want, but I can´t make sense of the use of ellipses - it seems to me that what you need are sections of parabolas, mirrored on both axis. The focus of the parabola will be adjusted to get the curve you want. I´m no math genious (that´s an understatement :)) but experimenting with ellipses they don´t seem to give equal path lengths - only approximations, whereas parabola sections will give exactly equal path lengths.

Sure, I´d like to try it out with a 3D printer - but it would be nice to hear what others think first, if it´s an idea worth pursuing or not.
 
Regarding your pile of junk dishes. Those are all offset types.
In other words, none of them contain "center" of the parabola.
There's a good chance you could clover 3 or four of them into
a much larger chunk of parabola without the leaves touching.

I had such a pile in my garage till recently...

Would be really hard to align for microwave purposes, which
was my original plan. But if you are just making a pair in the
back yard for acoustic toys....
 
Last edited:
Just thinking of a cross between paraline and Smith horn.
Straight internal reflectors at the top and bottom, rather
than the usual parabolic. Flat triangular expansion all the
way to the horizontal exit slot. Then two roundovers:
Halves of PVC pipe covering the entire frontal area.
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Yes, I have seen the video. It seems easy enough to make the curve you want, but I can´t make sense of the use of ellipses - it seems to me that what you need are sections of parabolas, mirrored on both axis. The focus of the parabola will be adjusted to get the curve you want. I´m no math genious (that´s an understatement :)) but experimenting with ellipses they don´t seem to give equal path lengths - only approximations, whereas parabola sections will give exactly equal path lengths.

The definition of an ellipse is that a segment from one focus to any point on ellipse to the other focal point is a constant along any point in ellipse. So in theory, the ellipse is exactly the shape you want if you want to use two drivers and one each at a focus. A parabola is a flat slice along a 3d ellipse surface of revolution.
 
Last edited:
Regarding your pile of junk dishes. Those are all offset types.
In other words, none of them contain "center" of the parabola.
There's a good chance you could clover 3 or four of them into
a much larger chunk of parabola without the leaves touching.

I had such a pile in my garage till recently...

Would be really hard to align for microwave purposes, which
was my original plan. But if you are just making a pair in the
back yard for acoustic toys....
Have you tried making a microwave transmission line?
Specific India made aluminium D channel works very well for proper e/h wave propagation. You cut the slots, cnc etc. Gains of 30dBi. I like coplanars when omni is a must. Easy to make, basically velocity factor vs freq, cut each segment, no more than ten (pairs, +&-) is best. Alternate connections between shield and center conductor. Connector and done. :)
 
Yeah, I shoved a sharpened 2x4 down a bunch of eggnog cans.
And using that as a form, hammered em' somewhat rectangular.
Getting them lined up well enough to solder, thats another story.
Trying to do one of those long slotted waveguide things, didn't
quite work as I had hoped, but probably my sloppy construction.

You mean coaxial co-linear? Those work great. I think a 2-hole
button between each segment works well to get things lined up.
I replaced the outer braid with brass tubing and run the center
wire down the outside of the brass for every other segment.
You still need to terminate the far end with a 1/4 wave stick
and the near end with 1/4 wave foldback balun. I put this all
in a loose fitting PEX pipe to protect it from collapse while also
minimize dielectric loading and losses.

Feed horns tried on the parabola were: a coffee can, and a patch,
and a quad patch, and a biquad with a reflector. And a very small
log periodic, and a 4 element Moxon cell Yagi.
 
Last edited:
Cassegrainalign???

I just figured how to maybe enhance my upcoming folded smithhorn.
Omit the optional parabola for now, not relevant to my immediate plan.
Bend the other way to match a narrow triangle to a wide triangle, thus
we are expanding faster all the way to the exit slot and roundovers.
 
It will all make much less sense if I just post a picture.
Pardon lazy drawing of multiple layers as-if only one...

Just for purpose of thinking too hard:
The speaker hole maybe 3inch, the panel 48wide x 24tall.
Since narrowing vertical directivity would be a feature,
the exit slot could be twice as wide as I drew it, originally
thinking 1/4 wave at 20KHz... Sonotube for roundovers?
 

Attachments

  • Smithagrain.jpg
    Smithagrain.jpg
    153.2 KB · Views: 449
Last edited: